Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Max retries exceeded with url: /v1/session #11

Open
LIUTAOGE opened this issue Jan 9, 2025 · 5 comments
Open

Max retries exceeded with url: /v1/session #11

LIUTAOGE opened this issue Jan 9, 2025 · 5 comments

Comments

@LIUTAOGE
Copy link

LIUTAOGE commented Jan 9, 2025

Image
Image
Continuous call will appear this error, how to solve

@SiriusNEO
Copy link
Collaborator

SiriusNEO commented Jan 9, 2025

Hi @LIUTAOGE , could you please provide a minimal reproduction for this problem? Thanks!

@LIUTAOGE
Copy link
Author

Image

when I use the main branch of code, I start serve also the sam error.
attn_func "xformers_fill_vllm_paged_attention_generate" or "xformers_fill_shared_prompts_generate" result to this error. Only when i start serve with the attn_func is "xformers_with_buffer", the serve is ok

why? Can i fix it with some param edit ?

@LIUTAOGE
Copy link
Author

Image

@LIUTAOGE
Copy link
Author

@SiriusNEO Thank you

@SiriusNEO
Copy link
Collaborator

@LIUTAOGE Hi! According to your description, this may due to incorrect installation of vLLM / parrot kernels. The ServeCore throws connection error since the engine is broken due to its internal error. You can try to modify and run the code under https://github.com/microsoft/ParrotServe/blob/main/benchmark/bench_kernel.py to check whether the kernel works.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants